xref: /petsc/src/ml/regressor/tests/output/ex2_prefix_tao_alt.out (revision eaa7cf455101e7391ead982166c6e99faf273403)
1PetscRegressor Object: 1 MPI process
2  type: linear
3    PetscRegressor Linear Type: ols
4    Intercept=0.0857143
5  Tao Object: (regressor_linear_) 1 MPI process
6    type: brgn
7      Regularizer weight: 0.
8      BRGN Regularization Type: l2prox
9      Tao Object: (tao_brgn_subsolver_) 1 MPI process
10        type: bnls
11          Tao Object: (tao_brgn_subsolver_tao_bnk_cg_) 1 MPI process
12            type: bncg
13              CG Type: ssml_bfgs
14              Skipped Stepdirection Updates: 0
15              Scaled gradient steps: 0
16              Pure gradient steps: 0
17              Not a descent direction: 0
18              Line search fails: 0
19              Matrix has not been preallocated yet
20            TaoLineSearch Object: (tao_brgn_subsolver_tao_bnk_cg_) 1 MPI process
21              type: more-thuente
22              maximum function evaluations=30
23              tolerances: ftol=0.0001, rtol=1e-10, gtol=0.9
24              total number of function evaluations=0
25              total number of gradient evaluations=0
26              total number of function/gradient evaluations=0
27              Termination reason: CONTINUE_ITERATING
28            convergence tolerances: gatol=1e-08,             grtol=1e-08,             steptol=0.,             gttol=0.
29            Residual in Function/Gradient:=0.
30            Objective value=0.
31            total number of iterations=0,                                    (max: 2000)
32            Solver never run
33          Rejected BFGS updates: 0
34          CG steps: 0
35          Newton steps: 1
36          BFGS steps: 0
37          Scaled gradient steps: 0
38          Gradient steps: 0
39          KSP termination reasons:
40            atol: 1
41            rtol: 0
42            ctol: 0
43            negc: 0
44            dtol: 0
45            iter: 0
46            othr: 0
47        TaoLineSearch Object: (tao_brgn_subsolver_) 1 MPI process
48          type: more-thuente
49          maximum function evaluations=30
50          tolerances: ftol=0.0001, rtol=1e-10, gtol=0.9
51          total number of function evaluations=0
52          total number of gradient evaluations=0
53          total number of function/gradient evaluations=1
54          Termination reason: SUCCESS
55        KSP Object: (tao_brgn_subsolver_tao_bnk_) 1 MPI process
56          type: stcg
57          maximum iterations=10000, initial guess is zero
58          tolerances: relative=1e-05, absolute=1e-50, divergence=10000.
59          left preconditioning
60          using UNPRECONDITIONED norm type for convergence test
61        PC Object: (tao_brgn_subsolver_tao_bnk_) 1 MPI process
62          type: lmvm
63          Mat Object: (tao_brgn_subsolver_tao_bnk_pc_lmvm_) 1 MPI process
64            type: lmvmbfgs
65            rows=2, cols=2
66              Max. storage: 5
67              Used storage: 0
68              Number of updates: 0
69              Number of rejects: 0
70              Number of resets: 0
71              J0:
72                Mat Object: (tao_brgn_subsolver_tao_bnk_pc_lmvm_mat_lmvm_J0_) 1 MPI process
73                  type: diagonal
74                  rows=2, cols=2
75                    Vec Object: 1 MPI process
76                      type: seq
77              Rescale type: diagonal
78              Rescale history: 1
79              Rescale params: alpha=1., beta=0.5, rho=1.
80              Rescale convex factor: theta=0.125
81          linear system matrix, which is also used to construct the preconditioner:
82          Mat Object: 1 MPI process
83            type: shell
84            rows=2, cols=2
85        total KSP iterations: 2
86        convergence tolerances: gatol=1e-08,         grtol=1e-08,         steptol=0.,         gttol=0.
87        Residual in Function/Gradient:=2.7894e-16
88        Objective value=0.0571429
89        total number of iterations=1,                                (max: 10000)
90        total number of function evaluations=3,                        (max: unlimited)
91        total number of function/gradient evaluations=3,            (max: unlimited)
92        total number of Hessian evaluations=1
93        total number of Jacobian evaluations=4
94        Solution converged:          ||g(X)|| <= gatol
95    convergence tolerances: gatol=1e-08,     grtol=1e-08,     steptol=0.,     gttol=0.
96    Residual in Function/Gradient:=0.
97    Objective value=1.15
98    total number of iterations=1,                            (max: 10000)
99    total number of function evaluations=3,                    (max: unlimited)
100    total number of function/gradient evaluations=3,        (max: unlimited)
101    total number of Hessian evaluations=1
102    Solution converged:      ||g(X)|| <= gatol
103Intercept is 0.085714
104Coefficients are
105Vec Object: 1 MPI process
106  type: seq
1070.4
1081.42857
109Predicted values are
110Vec Object: 1 MPI process
111  type: seq
1121.11429
1130.242857
1140.0857143
1150.642857
1161.91429
117